Original Title: Effect of PCI on Long-Term Survival in Patients with Stable Ischemic Heart Disease.
Reference: Steven P. Sedlis et al. N Engl J Med 2015;373:1937-46.

Courtesy of Dr. Santiago Alonso.
Centro Cardiológico Americano. Sanatorio Americano.
Montevideo, Uruguay.

During a median follow up of 4.6 years, with 2287 stable heart disease or silent ischemia patients, 1999 through 2004, the COURAGE trial had shown that even though PCI could relieve symptoms, it would not improve survival compared to optimal medical treatment (OMT) alone. However, survival curves tendency to separate in favor of PCI suggested a late benefit.

This study extended the COURAGE follow up to 1211 patients during 15 years, with a median follow up of 11.9 years (0 to 15.3).
Initially, there were 284 deaths (25%) in the PCI group and 277 (24%) in the OMT alone group (95% CI, 0.83 to 1.15; P=0.77), while the extended analysis showed 253 deaths (41%) in the PCI group vs. 253 (42%) in the OMT alone group (95% CI, 0.7 to 1.13; P=0.53).

Conclusion:
In patients with stable heart disease or silent ischemia there were no differences in survival at long term between those treated with PCI in addition to medical treatment vs. those receiving medical treatment alone.

Editorial Comment:
This extended analysis at 15 years supports the original study outcomes. The higher mortality in the extended follow up subgroup can be explained not only by its natural evolution (especially because it analyzed all-cause mortality), but also by the higher risk profile of this last subgroup. Other limitations to note are: DES were included only in the last 6 months of enrollment and only 3% of the population saw the benefit of these devices. In addition, FFR and IVUS were rarely used.

32.6% of patients crossed over to myocardial revascularization surgery during the original follow up with a crossover rate after a year of 2.7% per year, and there are no data on cross overs during the extended follow up. Therefore, we lack the evidence to arrive at valid conclusions on the kind of treatment the original patients received. Finally, the extended analysis included only 53% of the original population.

It would be interesting to obtain outcomes from a similar study with a long term follow up using the latest technology, both as regards devices and medical treatment, and clearly account for crossover patients during the whole evolution.